Ischemic Brain Metabolism in Patients With Chronic Cerebrovascular Disease: Increased Oxygen Extraction Fraction and Cerebrospinal Fluid Lactate Setsuro Ibayashi, MD, PhD, Katsumi Irie, MD, PhD, Jiro Kitayama, MD, Tetsuhiko Nagao, MD, PhD, Takanari Kitazono, MD, PhD, and Masatoshi Fujishima, MD, PhD The aim of the present study is to elucidate the existence of chronically ischemic metabolism concomitant with misery perfusion of the brain in patients with chronic cerebrovascular disease. For this purpose, we measured cerebral blood flow (CBF) and oxygen metabolism by positron emission tomography (PET) and also determined cerebrospinal fluid (CSF) lactate as an indicator of the ischemic brain metabolism. Twenty-eight patients with chronic ischemic stroke and transient ischemic attack (TIA), who had angiographically occlusive (n ⫽ 11), stenotic (n ⫽ 10), and nonstenotic changes (n ⫽ 7) of the carotid artery and/or the intracranial major artery, were selected for this study. CBF, oxygen extraction fraction (OEF), cerebral metabolic rate for oxygen (CMRO2), and cerebral blood volume (CBV) were determined by PET, and CSF lactate and pyruvate were determined by enzymatic method in the patients with various grades of stenotic changes of the carotid artery. There were no significant differences in PET parameters and CSF variables among the groups of the occlusive, stenotic, and nonstenotic carotid artery. However, CSF lactate was correlated negatively with mean bilateral hemispheric (m)CBF (R2 ⫽ 0.229, P ⬍ .01), positively with mOEF (R2 ⫽ 0.278, P ⬍ .005) and more highly with mCMRO2/CBF (absolute extraction of oxygen content to the brain) (R2 ⫽ 0.473, P ⬍ .0001) in all patients. There was no correlation between CSF lactate and mCMRO2 or mCBV. None of the cases in the nonstenotic group showed mOEF greater than 0.45, or mCMRO2/CBF greater than 7.9 vol%, while 80% of the cases in the stenotic group and 82% of the cases in the occlusive group showed mOEF and mCMRO2/CBF exceeding the above-mentioned values, respectively. The present findings, that increased mOEF and mCMRO2/CBF were significantly correlated with increased CSF lactate, indicate the brain to be in a metabolically ischemic state or increased anaerobic glycolysis with oxygen metabolism maintained in patients with chronic ischemic stroke. Key Words: Misery perfusion—Oxygen extraction fraction—Cerebrospinal fluid lactate—Ischemic metabolism.
Pathophysiology of ischemic stroke has been investigated hemodynamically by using single photon emission computed tomography (SPECT) and positron emission
From the Second Department of Internal Medicine, Faculty of Medicine, Kyushu University, Higashi-ku, Fukuoka, Japan. Received December 22, 1999; accepted January 7, 2000. Address reprint requests to Setsuro Ibayashi, MD, PhD, Second Department of Internal Medicine, Faculty of Medicine, Kyushu University, Maidashi 3-1-1, Higashi-ku, Fukuoka, 812-8582 Japan. Copyright r 2000 by National Stroke Association 1052-3057/00/0904-0003$3.00/0 doi:10.1053/jscd.2000.7219
166
tomography (PET)1-3 and morphologically by brain imaging. In the acute stage of brain ischemia, the ischemic core shows minimal to very low blood flow with metabolic derangement, resulting in ischemic neuronal death, while its surrounding area shows hypoperfusion or misery perfusion. It has been known in chronic brain infarction that cerebral blood flow (CBF) and cerebral oxygen metabolism are greatly decreased, but not null in the infarcted area, and also less greatly decreased in its surrounding regions and even in the remote area.4,5 Such flow and metabolism reductions in chronic stage of infarction provide 2 possibilities for explanation. One
Journal of Stroke and Cerebrovascular Diseases, Vol. 9, No. 4 ( July-August), 2000: pp 166-171
ISCHEMIC BRAIN METABOLISM IN CHRONIC CEREBROVASCULAR DISEASE
possibility is that the reduced tissue metabolism due to decreased demand is a primary phenomenon, therefore, leading to the consequent reduction of CBF (flowmetabolism coupling or matched hypoperfusion).3,6,7 Another possibility is that hypoperfusion or insufficient blood supply to the brain per se is a primary phenomenon, which is due to stenotic or occlusive lesions of the supplying artery to the brain, resulting in decreased brain metabolism (flow-metabolism uncoupling or misery perfusion).8 In the former, CBF is decreased proportionally as cerebral oxygen or glucose metabolism is reduced. Therefore, cerebral oxygen extraction remains unchanged, indicating matched reduction of flow-metabolism. As a result, ischemic brain metabolite, such as tissue lactate, does not seem to increase in this situation. In the latter, persistent hypoperfusion of the brain, which is accompanied by the increased oxygen extraction fraction (OEF) for compensatory mechanism to maintain brain metabolism, may lead to ischemic metabolism, therefore, metabolites of anaerobic glycolysis could be increased. The aim of the present study is to elucidate the existence of chronically ischemic metabolism concomitant with misery perfusion of the brain in patients with chronic cerebrovascular disease. For this purpose, we measured CBF, cerebral oxygen uptake, and OEF by PET and also determined lactate and pyruvate concentrations in the cerebrospinal fluid (CSF) as an indicator of the brain tissue anaerobic metabolism in the ischemic stroke patients with various grades of stenotic changes of the carotid artery.
Materials and Methods Subjects Twenty-eight in-patients with chronic ischemic stroke, who underwent both PET and CSF study from February 1985 to March 1995, were selected for this study. No patients were included in the present study who had traumatic brain infarction or brain infarction of cardiogenic embolism. Of these, 22 patients had atherothrombotic or lacunar brain infarction and 6 patients had transient ischemic attacks (TIA). Cerebral angiography in addition to ultrasonography or magnetic resonance (MR) angiography was performed to determine the extra- and intracranial arterial lesions in all patients except for 1 who had undergone an echographical determination alone. Eleven patients had unilateral carotid artery occlusion at a bifurcation level, 10 patients showed mild to moderate stenotic lesions of carotid or major cerebral arteries (stenosis of 25% to 75% by North American Symptomatic Carotid Endarterectomy Trial [NASCET] criteria9 on angiography), and the remaining 7 patients had no significant but slight stenosis less than 25% or small irregularity of the intra- and extracranial arteries.
167
PET Study Informed consent was obtained from all patients before the PET study according to the approval by the ethics committee of Kyushu University Hospital. Regional cerebral blood flow (rCBF) and regional oxygen extraction fraction (rOEF) were measured by the H215O continuous infusion method and the 15O2 continuous inhalation method, respectively, according to the oxygen-15 steadystate techniques.10 Regional cerebral metabolic rate for oxygen (rCMRO2) was calculated as rCMRO2 ⫽ rCBF ⫻ rOEF ⫻ arterial oxygen content. Both rOEF and rCMRO2 were corrected using regional cerebral blood volume (rCBV) measured using a single inhalation of C15O gas. We used a HEADTOME-III device (Shimadzu Inc, Kyoto, Japan and Akita Noken, Akita, Japan) with planar and axial resolutions of 8.2 mm full-width at half-maximum and corrected for attenuation by means of a transmission scan using an external germanium-68-gallium-68 ring source in each patient. During the PET study, each patient was placed in a supine position with the eyes open and the ears unplugged in a dimly lit room. One femoral artery was cannulated to sample blood for radioactive concentration and blood gases. Arterial blood gases, as well as hemoglobin, hematocrit, and blood glucose levels, were analyzed at the start and end of each scan. Regions of interest 18 ⫻ 14 mm were set on the bilateral cerebral cortical area (frontal, temporal, parietal, and occipital cortices), white matter (centrum semiovale) and deep gray matter (striatum and thalamus), cerebellar hemisphere, and the brainstem using the brain computed tomography (CT) image corresponding to each PET slice as a reference. In addition, regions of interest were set on whole hemispheric areas on the PET slice 50 mm above the orbitomeatal line where the values for PET parameters were used as the bilateral hemispheric mean values such as mCBF, mOEF, and mCMRO2. Mean cerebral extraction of oxygen content (mCMRO2/CBF) was calculated by mCMRO2/mCBF, which was equal to cerebral arteriovenous difference of oxygen content or absolute extraction of oxygen content from the blood to the brain. The PET data analyses were performed by 2 of the authors (K.I. and J.K.) who were blinded to the CSF parameters. CSF Study Lumbar puncture was performed after explaining the medical significance and purpose of our clinical study and obtaining informed consent from all patients or their family if needed. We also checked that the patients had no papilledema and intracranial hemorrhagic transformation. All CSF samples were clear and showed normal cell counts (less than 5/µL). Immediately after sampling, 1.5 mL of CSF was added to the same amount of 10% perchloric acid to remove protein, and the supernatant was stored in a deep freezer until estimation. After neutralization with potassium hydroxide, lactate and
S. IBAYASHI ET AL.
168
Table 1. Mean values for PET parameters and CSF chemistry in chronic ischemic cerebrovascular patients with various steno-occlusive lesions of carotid and intracranial major arteries PET
Vascular lesion
No. of patients
Occlusion Stenosis None
11 10 7
Age (yr)
mCMRO2 mCBF (mL/100 g/ (mL/100 g/ min) min)
mOEF
mCMRO2/ CBF (O2 vol%)
CSF mCBV (mL/100 g)
Lactate (mmol/L)
Pyruvate (mmol/L)
L/P
54 ⫾ 4 27.4 ⫾ 1.6 2.02 ⫾ 0.11 0.44 ⫾ 0.02 7.51 ⫾ 0.42 4.55 ⫾ 0.39 1.74 ⫾ 0.07 0.12 ⫾ 0.01 14.6 ⫾ 0.4 63 ⫾ 2 30.9 ⫾ 2.4 2.20 ⫾ 0.16 0.43 ⫾ 0.02 7.34 ⫾ 0.51 4.29 ⫾ 0.21 1.75 ⫾ 0.10 0.18 ⫾ 0.07 14.3 ⫾ 1.7 56 ⫾ 3 28.7 ⫾ 1.7 1.97 ⫾ 0.12 0.39 ⫾ 0.02 6.91 ⫾ 0.30 3.85 ⫾ 0.24 1.65 ⫾ 0.05 0.11 ⫾ 0.01 15.1 ⫾ 0.5
NOTE: Values are mean ⫾ SEM.
pyruvate in CSF were analyzed by standard enzymatic methods (Boehringer; Mannheim, Germany).11 CSF protein and glucose concentrations were also determined. Determination The PET study was performed first, followed by the CSF study in 16 patients, and vice versa in 12 patients. The interval between the PET study and the CSF sampling was 17.4 ⫾ 6.2 standard error of mean (SEM) days. The interval between the onset of stroke or TIA and PET study was 2.7 ⫾ 0.4 months, which is a chronic stage. Statistical Analysis All data are given as mean ⫾ SEM. We compared mean values for PET parameters and CSF variables among the patients with occlusive, stenotic and nonstenotic lesions of the carotid and/or intracranial major arteries using analysis of variance (ANOVA) and Fisher’s probable least significant difference (PLSD). The relationship between CSF lactate and mCBF, mCMRO2, mOEF, mCBV or mCMRO2/CBF was evaluated by linear regression analysis. A P value of less than .05 was considered statistically significant.
Results Parameters for PET and CSF Study There were no differences in hemoglobin, hematocrit, blood glucose levels, and arterial gas parameters among the groups in the study (data not shown). Mean values for PET parameters and CSF lactate, pyruvate, and lactate/ pyruvate (L/P) ratio in the groups of patients with occlusive, stenotic, and nonstenotic lesions are depicted in Table 1. There were no significant differences in age, mCBF, mCMRO2, mOEF, mCMRO2/CBF, and mCBV among the 3 groups, although mOEF and mCMRO2/CBF in the occlusion group tended to be higher than those in other groups. CSF lactate was higher in occlusive and stenotic groups than in the nonstenotic group, however, its difference was not significant. CSF L/P ratio did not differ among the groups, or did CSF protein and glucose levels (data not shown). Relationship of CSF Lactate and PET Parameters There was a significant negative correlation between CSF lactate and mCBF in all patients (P ⬍ .01) (Fig 1), indicating that the lowering of mCBF causes an increase
Figure 1. A significant negative correlation of CSF lactate with mean cerebral blood flow (mCBF) in all patients with chronic ischemic cerebrovascular disease. Angiography showed occlusion (䊉), mild to moderate stenosis (䉱), and slight or no stenosis (䊊) of the carotid artery or intracranial major arteries.
ISCHEMIC BRAIN METABOLISM IN CHRONIC CEREBROVASCULAR DISEASE
169
Figure 2. Lack of relationship between CSF lactate and mean cerebral metabolic rate for oxygen (mCMRO2) in all patients with chronic ischemic cerebrovascular disease.
in CSF lactate, which implies the increased anaerobic metabolite in the brain. There was no relationship between CSF lactate and mCMRO2 (Fig 2), indicating that the lowered cerebral oxygen utilization per se is not accompanied by increased CSF lactate. No significant correlation was found between CSF lactate and mCBV. In contrast, CSF lactate was linearly correlated with mOEF (P ⬍ .005) (Fig 3), and also highly correlated with mCMRO2/CBF in all patients (P ⬍ .0001) (Fig 4). Correlation coefficiency in the latter was much greater (R2 ⫽ 0.473) than that in the former (R2 ⫽ 0.278). None of the patients in the nonstenotic group showed mOEF greater than 0.45, or mCMRO2/CBF greater than 7.9 vol%, while 8 patients (80%) in the stenotic group and 9 patients (82%) in the occlusive group showed values exceeding 0.45 for mOEF and 7.9 vol% for mCMRO2/CBF, respectively.
Discussion The present study is the first to describe in chronic ischemic stroke or TIA that lowered CBF and either increased OEF or increased cerebral oxygen content extraction calculated from mCMRO2/CBF are associated with increased CSF lactate, indicating the existence of chronically ischemic state or increased anaerobic metabolism in
the brain. There have been many observations showing low CBF with increased OEF in chronic stroke patients or in those who have internal carotid stenosis or occlusion.5,8,12,13 Misery perfusion with increased OEF was found in 18.3% of brain infarction of middle cerebral artery territory with aphasia.5 In those, 7.5% of the patients were in the chronic stage of stroke, ranged poststroke 30 and 250 days. Furthermore, increased OEF was observed not only in the infarcted hemisphere, but also in the contralateral or noninfarcted areas or even in the remote areas, mostly due to hemodynamic factors. They include lowered cerebral perfusion pressure due to steno-occlusive lesions of the supplying artery, resulting in insufficient blood flow and oxygen supply to the tissue, and insufficient oxygen extraction from circulating blood to the brain despite maximally increased OEF by compensatory mechanism in response to reduced CBF. There have been reports describing marked increase in CSF lactate and L/P ratio in acute ischemic stroke, followed by a gradual fall in chronic stage.14,15 However, high CSF lactate in chronic ischemic stroke has been experienced in our laboratory. Increased CSF lactate, resulting from increased tissue lactate, could be due possibly to anaerobic metabolism in infarcted brain, and also in noninfarcted, but low perfused, areas of the brain. As shown in Figs 3 and 4, increased mOEF and mCMRO2/
Figure 3. A significant linear correlation of CSF lactate and mOEF in all patients with chronic ischemic cerebrovascular disease.
S. IBAYASHI ET AL.
170
Figure 4. A highly significant correlation of CSF lactate and mean cerebral metabolic rate for oxygen divided by mean cerebral blood flow (mCMRO2/CBF), which indicates absolute value for oxygen content extraction from the blood to the brain, in all patients with chronic ischemic cerebrovascular disease.
CBF beyond a certain level (0.45% and 7.9 vol%, respectively, in this study) suggest the possible existence of the steno-occlusive lesions of the supplying artery, which may lead to increased oxygen extraction to the brain because of insufficient flow. However, increased oxygen extraction of compensatory mechanism for reduced blood supply is not sufficient enough to keep normal brain metabolism, and, therefore, the brain is in enhanced anaerobic glycolysis. It is possible that the increased CSF lactate is an epiphenomenon resulting from gliosis and macrophage accumulation in patients in subacute to chronic infarcts. Finally, lactate is increased in the brain, as well as in CSF. These results strongly suggest the existence of a metabolically ischemic state of the brain (increased anaerobic glycolysis with oxygen metabolism maintained) in chronic cerebral infarction or TIA, particularly in some patients who have steno-occlusive vascular lesions. Misery perfusion or mismatch between CBF and CMRO2 has been reported in patients with occluded internal carotid artery, in which up to 20% of patients may experience delayed ischemic events in the ipsilateral cerebral hemisphere. Baron et al4 described in stroke patients with carotid occlusion less decrease in cerebral metabolic rate for glucose than that for oxygen (CMRglc ⬎ CMRO2), indicating enhanced anaerobic glycolysis. Their findings support our results of enhanced anaerobic glycolysis and increased CSF lactate. Correlation coefficiency was much greater between CSF lactate and mCMRO2/CBF than that between CSF lactate and mOEF, indicating that the value for mCMRO2/CBF is related to, but not equal to, mOEF. In an anemic condition, the difference between 2 parameters becomes greater than that in a nonanemic condition, because CMRO2/CBF is the same as OEF ⫻ arterial O2 content. Therefore, absolute value and fraction rate of cerebral oxygen extraction might be different, depending on the hemoglobin level and oxygen saturation of arterial blood. As mentioned in the Results, there were no essential differences in hemoglobin, hematocrit, and arterial gas parameters including oxygen saturation among the groups in our study.
The present results are strongly suggestive in aspect to hemodynamic treatment in patients with chronic brain infarction with increased OEF greater than 0.45 or increased mCMRO2/CBF greater than 7.9 vol% of oxygen, in which the brain becomes metabolically ischemic or anaerobic. However, CMRO2 is not indicative of ischemic metabolism of the brain. In those patients, restoration of the reduced cerebral perfusion pressure might be promising against the stroke recurrence either by medical treatment with vasodilators or with raising systemic blood pressure,16 or by surgical treatment with endarterectomy of steno-occlusive vessels or extraintracranial bypass.12,13 To confirm our findings more specifically, simultaneous determinations of rOEF and rCMRO2/rCBF by PET and regional tissue lactate or adenosine triphosphate by 1H- or 32P magnetic resonance spectroscopy (MRS) are needed to show a direct correlation of increased OEF and lactate. Irie et al17 from our laboratory have reported in multiinfarct dementia that CSF lactate slightly, but significantly increased, which was well correlated with decreased CBF and increased OEF, but not with CMRO2, indicating that cerebral ischemia with impaired aerobic glycolysis plays a role in dementia in multiple infarction. Therefore, we must study the cognitive function in chronic ischemic stroke patients with steno-occlusive changes in carotid or intracranial major arteries. In conclusion, increased OEF and CMRO2/CBF were significantly correlated with increased CSF lactate, indicating the brain to be in a metabolically ischemic state or in enhanced anaerobic glycolysis with oxygen metabolism maintained in patients with chronic ischemic stroke.
References 1. Powers WJ. Cerebral hemodynamics in ischemic cerebrovascular disease. Ann Neurol 1991;29:231-240. 2. Heiss W-D. Application of positron emission tomography to the study of cerebral ischemia. In: Ginsberg MD, Bogousslavsky J, eds. Cerebrovascular disease: Pathophysiology, diagnosis, and management. Malden: Blackwell Science, 1998;761-772.
ISCHEMIC BRAIN METABOLISM IN CHRONIC CEREBROVASCULAR DISEASE 3. Frackowiak RSJ, Pozzilli C, Legg NJ, et al. Regional cerebral oxygen supply and utilization in dementia. A clinical and physiological study with oxygen-15 positron tomography. Brain 1981;104:753-778. 4. Baron JC, Bousser MG, Rey A, et al. Reversal of focal ‘‘misery-perfusion syndrome’’ by extra-intracranial arterial bypass in hemodynamic cerebral ischemia. A case study with 15O positron emission tomography. Stroke 1981;12:454-459. 5. Tsutsumi K, Nagata K. Misery perfusion syndrome at the chronic stage of cerebral infarction (in Japanese with English abstract). Jpn J Stroke 1998;20:489-499. 6. Lenzi GL, Franckowiak RSJ, Jones T. Cerebral oxygen metabolism and blood flow in human cerebral ischemic infarction. J Cereb Blood Flow Metab 1982;2:321-335. 7. Gibbs JM, Frackowiak RSJ, Legg NJ. Regional cerebral blood flow and oxygen metabolism in dementia due to vascular disease. Gerontology 1986;32:84-88, (suppl 1). 8. Yamauchi H, Fukuyama H, Fujimoto N, et al. Significance of low perfusion with increased oxygen extraction fraction in a case of internal carotid artery stenosis. Stroke 1992;23:431-432. 9. North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med 1991;325:445-453. 10. Fujii K, Sadoshima S, Okada Y, et al. Cerebral blood flow and metabolism in normotensive and hypertensive pa-
11.
12.
13.
14.
15.
16.
17.
171
tients with transient neurologic deficits. Stroke 1990;21: 283-290. Fujishima M, Sugi T, Choki J, et al. Cerebrospinal fluid and arterial lactate, pyruvate and acid-base balance in patients with intracranial hemorrhages. Stroke 1975;6:707714. Gibbs JM, Wise RJS, Leenders KL, et al. Evaluation of cerebral perfusion reserve in patients with carotid-artery occlusion. Lancet 1984;1:310-314. Busse O, Hoffmann O. CSF lactate and CT findings in middle cerebral artery infarction. A comparative study. Stroke 1983;14:960-963. Berger JP, Frawer R. Cerebrospinal fluid (CSF) lactate and pyruvate in acute neurological situations. In: Bossart H, Lausanne P, eds. Lactate in acute conditions. Basel: Karger, 1979:115-133. Baron JC, Rougemont D, Soussaline F, et al. Local interrelationships of cerebral oxygen consumption and glucose utilization in normal subjects and in ischemic stroke patients: A positron tomography study. J Cereb Blood Flow Metab 1984;4:140-149. Meyer JS, Judd B, Tawaklna T, et al. Improved cognition after control of risk factors for multi-infarct dementia. JAMA 1986;256:2203-2209. Irie K, Ibayashi S, Fujii K, et al. Relationship between cerebral blood flow, metabolism and cerebrospinal fluid lactate in multi-infarct dementia (in Japanese with English abstract). Jpn J Stroke 1994;16:15-20.
Pathophysiology of ischemic stroke has been investigated hemodynamically by using single photon emission computed tomography (SPECT) and positron emission
From the Second Department of Internal Medicine, Faculty of Medicine, Kyushu University, Higashi-ku, Fukuoka, Japan. Received December 22, 1999; accepted January 7, 2000. Address reprint requests to Setsuro Ibayashi, MD, PhD, Second Department of Internal Medicine, Faculty of Medicine, Kyushu University, Maidashi 3-1-1, Higashi-ku, Fukuoka, 812-8582 Japan. Copyright r 2000 by National Stroke Association 1052-3057/00/0904-0003$3.00/0 doi:10.1053/jscd.2000.7219
166
tomography (PET)1-3 and morphologically by brain imaging. In the acute stage of brain ischemia, the ischemic core shows minimal to very low blood flow with metabolic derangement, resulting in ischemic neuronal death, while its surrounding area shows hypoperfusion or misery perfusion. It has been known in chronic brain infarction that cerebral blood flow (CBF) and cerebral oxygen metabolism are greatly decreased, but not null in the infarcted area, and also less greatly decreased in its surrounding regions and even in the remote area.4,5 Such flow and metabolism reductions in chronic stage of infarction provide 2 possibilities for explanation. One
Journal of Stroke and Cerebrovascular Diseases, Vol. 9, No. 4 ( July-August), 2000: pp 166-171
ISCHEMIC BRAIN METABOLISM IN CHRONIC CEREBROVASCULAR DISEASE
possibility is that the reduced tissue metabolism due to decreased demand is a primary phenomenon, therefore, leading to the consequent reduction of CBF (flowmetabolism coupling or matched hypoperfusion).3,6,7 Another possibility is that hypoperfusion or insufficient blood supply to the brain per se is a primary phenomenon, which is due to stenotic or occlusive lesions of the supplying artery to the brain, resulting in decreased brain metabolism (flow-metabolism uncoupling or misery perfusion).8 In the former, CBF is decreased proportionally as cerebral oxygen or glucose metabolism is reduced. Therefore, cerebral oxygen extraction remains unchanged, indicating matched reduction of flow-metabolism. As a result, ischemic brain metabolite, such as tissue lactate, does not seem to increase in this situation. In the latter, persistent hypoperfusion of the brain, which is accompanied by the increased oxygen extraction fraction (OEF) for compensatory mechanism to maintain brain metabolism, may lead to ischemic metabolism, therefore, metabolites of anaerobic glycolysis could be increased. The aim of the present study is to elucidate the existence of chronically ischemic metabolism concomitant with misery perfusion of the brain in patients with chronic cerebrovascular disease. For this purpose, we measured CBF, cerebral oxygen uptake, and OEF by PET and also determined lactate and pyruvate concentrations in the cerebrospinal fluid (CSF) as an indicator of the brain tissue anaerobic metabolism in the ischemic stroke patients with various grades of stenotic changes of the carotid artery.
Materials and Methods Subjects Twenty-eight in-patients with chronic ischemic stroke, who underwent both PET and CSF study from February 1985 to March 1995, were selected for this study. No patients were included in the present study who had traumatic brain infarction or brain infarction of cardiogenic embolism. Of these, 22 patients had atherothrombotic or lacunar brain infarction and 6 patients had transient ischemic attacks (TIA). Cerebral angiography in addition to ultrasonography or magnetic resonance (MR) angiography was performed to determine the extra- and intracranial arterial lesions in all patients except for 1 who had undergone an echographical determination alone. Eleven patients had unilateral carotid artery occlusion at a bifurcation level, 10 patients showed mild to moderate stenotic lesions of carotid or major cerebral arteries (stenosis of 25% to 75% by North American Symptomatic Carotid Endarterectomy Trial [NASCET] criteria9 on angiography), and the remaining 7 patients had no significant but slight stenosis less than 25% or small irregularity of the intra- and extracranial arteries.
167
PET Study Informed consent was obtained from all patients before the PET study according to the approval by the ethics committee of Kyushu University Hospital. Regional cerebral blood flow (rCBF) and regional oxygen extraction fraction (rOEF) were measured by the H215O continuous infusion method and the 15O2 continuous inhalation method, respectively, according to the oxygen-15 steadystate techniques.10 Regional cerebral metabolic rate for oxygen (rCMRO2) was calculated as rCMRO2 ⫽ rCBF ⫻ rOEF ⫻ arterial oxygen content. Both rOEF and rCMRO2 were corrected using regional cerebral blood volume (rCBV) measured using a single inhalation of C15O gas. We used a HEADTOME-III device (Shimadzu Inc, Kyoto, Japan and Akita Noken, Akita, Japan) with planar and axial resolutions of 8.2 mm full-width at half-maximum and corrected for attenuation by means of a transmission scan using an external germanium-68-gallium-68 ring source in each patient. During the PET study, each patient was placed in a supine position with the eyes open and the ears unplugged in a dimly lit room. One femoral artery was cannulated to sample blood for radioactive concentration and blood gases. Arterial blood gases, as well as hemoglobin, hematocrit, and blood glucose levels, were analyzed at the start and end of each scan. Regions of interest 18 ⫻ 14 mm were set on the bilateral cerebral cortical area (frontal, temporal, parietal, and occipital cortices), white matter (centrum semiovale) and deep gray matter (striatum and thalamus), cerebellar hemisphere, and the brainstem using the brain computed tomography (CT) image corresponding to each PET slice as a reference. In addition, regions of interest were set on whole hemispheric areas on the PET slice 50 mm above the orbitomeatal line where the values for PET parameters were used as the bilateral hemispheric mean values such as mCBF, mOEF, and mCMRO2. Mean cerebral extraction of oxygen content (mCMRO2/CBF) was calculated by mCMRO2/mCBF, which was equal to cerebral arteriovenous difference of oxygen content or absolute extraction of oxygen content from the blood to the brain. The PET data analyses were performed by 2 of the authors (K.I. and J.K.) who were blinded to the CSF parameters. CSF Study Lumbar puncture was performed after explaining the medical significance and purpose of our clinical study and obtaining informed consent from all patients or their family if needed. We also checked that the patients had no papilledema and intracranial hemorrhagic transformation. All CSF samples were clear and showed normal cell counts (less than 5/µL). Immediately after sampling, 1.5 mL of CSF was added to the same amount of 10% perchloric acid to remove protein, and the supernatant was stored in a deep freezer until estimation. After neutralization with potassium hydroxide, lactate and
S. IBAYASHI ET AL.
168
Table 1. Mean values for PET parameters and CSF chemistry in chronic ischemic cerebrovascular patients with various steno-occlusive lesions of carotid and intracranial major arteries PET
Vascular lesion
No. of patients
Occlusion Stenosis None
11 10 7
Age (yr)
mCMRO2 mCBF (mL/100 g/ (mL/100 g/ min) min)
mOEF
mCMRO2/ CBF (O2 vol%)
CSF mCBV (mL/100 g)
Lactate (mmol/L)
Pyruvate (mmol/L)
L/P
54 ⫾ 4 27.4 ⫾ 1.6 2.02 ⫾ 0.11 0.44 ⫾ 0.02 7.51 ⫾ 0.42 4.55 ⫾ 0.39 1.74 ⫾ 0.07 0.12 ⫾ 0.01 14.6 ⫾ 0.4 63 ⫾ 2 30.9 ⫾ 2.4 2.20 ⫾ 0.16 0.43 ⫾ 0.02 7.34 ⫾ 0.51 4.29 ⫾ 0.21 1.75 ⫾ 0.10 0.18 ⫾ 0.07 14.3 ⫾ 1.7 56 ⫾ 3 28.7 ⫾ 1.7 1.97 ⫾ 0.12 0.39 ⫾ 0.02 6.91 ⫾ 0.30 3.85 ⫾ 0.24 1.65 ⫾ 0.05 0.11 ⫾ 0.01 15.1 ⫾ 0.5
NOTE: Values are mean ⫾ SEM.
pyruvate in CSF were analyzed by standard enzymatic methods (Boehringer; Mannheim, Germany).11 CSF protein and glucose concentrations were also determined. Determination The PET study was performed first, followed by the CSF study in 16 patients, and vice versa in 12 patients. The interval between the PET study and the CSF sampling was 17.4 ⫾ 6.2 standard error of mean (SEM) days. The interval between the onset of stroke or TIA and PET study was 2.7 ⫾ 0.4 months, which is a chronic stage. Statistical Analysis All data are given as mean ⫾ SEM. We compared mean values for PET parameters and CSF variables among the patients with occlusive, stenotic and nonstenotic lesions of the carotid and/or intracranial major arteries using analysis of variance (ANOVA) and Fisher’s probable least significant difference (PLSD). The relationship between CSF lactate and mCBF, mCMRO2, mOEF, mCBV or mCMRO2/CBF was evaluated by linear regression analysis. A P value of less than .05 was considered statistically significant.
Results Parameters for PET and CSF Study There were no differences in hemoglobin, hematocrit, blood glucose levels, and arterial gas parameters among the groups in the study (data not shown). Mean values for PET parameters and CSF lactate, pyruvate, and lactate/ pyruvate (L/P) ratio in the groups of patients with occlusive, stenotic, and nonstenotic lesions are depicted in Table 1. There were no significant differences in age, mCBF, mCMRO2, mOEF, mCMRO2/CBF, and mCBV among the 3 groups, although mOEF and mCMRO2/CBF in the occlusion group tended to be higher than those in other groups. CSF lactate was higher in occlusive and stenotic groups than in the nonstenotic group, however, its difference was not significant. CSF L/P ratio did not differ among the groups, or did CSF protein and glucose levels (data not shown). Relationship of CSF Lactate and PET Parameters There was a significant negative correlation between CSF lactate and mCBF in all patients (P ⬍ .01) (Fig 1), indicating that the lowering of mCBF causes an increase
Figure 1. A significant negative correlation of CSF lactate with mean cerebral blood flow (mCBF) in all patients with chronic ischemic cerebrovascular disease. Angiography showed occlusion (䊉), mild to moderate stenosis (䉱), and slight or no stenosis (䊊) of the carotid artery or intracranial major arteries.
ISCHEMIC BRAIN METABOLISM IN CHRONIC CEREBROVASCULAR DISEASE
169
Figure 2. Lack of relationship between CSF lactate and mean cerebral metabolic rate for oxygen (mCMRO2) in all patients with chronic ischemic cerebrovascular disease.
in CSF lactate, which implies the increased anaerobic metabolite in the brain. There was no relationship between CSF lactate and mCMRO2 (Fig 2), indicating that the lowered cerebral oxygen utilization per se is not accompanied by increased CSF lactate. No significant correlation was found between CSF lactate and mCBV. In contrast, CSF lactate was linearly correlated with mOEF (P ⬍ .005) (Fig 3), and also highly correlated with mCMRO2/CBF in all patients (P ⬍ .0001) (Fig 4). Correlation coefficiency in the latter was much greater (R2 ⫽ 0.473) than that in the former (R2 ⫽ 0.278). None of the patients in the nonstenotic group showed mOEF greater than 0.45, or mCMRO2/CBF greater than 7.9 vol%, while 8 patients (80%) in the stenotic group and 9 patients (82%) in the occlusive group showed values exceeding 0.45 for mOEF and 7.9 vol% for mCMRO2/CBF, respectively.
Discussion The present study is the first to describe in chronic ischemic stroke or TIA that lowered CBF and either increased OEF or increased cerebral oxygen content extraction calculated from mCMRO2/CBF are associated with increased CSF lactate, indicating the existence of chronically ischemic state or increased anaerobic metabolism in
the brain. There have been many observations showing low CBF with increased OEF in chronic stroke patients or in those who have internal carotid stenosis or occlusion.5,8,12,13 Misery perfusion with increased OEF was found in 18.3% of brain infarction of middle cerebral artery territory with aphasia.5 In those, 7.5% of the patients were in the chronic stage of stroke, ranged poststroke 30 and 250 days. Furthermore, increased OEF was observed not only in the infarcted hemisphere, but also in the contralateral or noninfarcted areas or even in the remote areas, mostly due to hemodynamic factors. They include lowered cerebral perfusion pressure due to steno-occlusive lesions of the supplying artery, resulting in insufficient blood flow and oxygen supply to the tissue, and insufficient oxygen extraction from circulating blood to the brain despite maximally increased OEF by compensatory mechanism in response to reduced CBF. There have been reports describing marked increase in CSF lactate and L/P ratio in acute ischemic stroke, followed by a gradual fall in chronic stage.14,15 However, high CSF lactate in chronic ischemic stroke has been experienced in our laboratory. Increased CSF lactate, resulting from increased tissue lactate, could be due possibly to anaerobic metabolism in infarcted brain, and also in noninfarcted, but low perfused, areas of the brain. As shown in Figs 3 and 4, increased mOEF and mCMRO2/
Figure 3. A significant linear correlation of CSF lactate and mOEF in all patients with chronic ischemic cerebrovascular disease.
S. IBAYASHI ET AL.
170
Figure 4. A highly significant correlation of CSF lactate and mean cerebral metabolic rate for oxygen divided by mean cerebral blood flow (mCMRO2/CBF), which indicates absolute value for oxygen content extraction from the blood to the brain, in all patients with chronic ischemic cerebrovascular disease.
CBF beyond a certain level (0.45% and 7.9 vol%, respectively, in this study) suggest the possible existence of the steno-occlusive lesions of the supplying artery, which may lead to increased oxygen extraction to the brain because of insufficient flow. However, increased oxygen extraction of compensatory mechanism for reduced blood supply is not sufficient enough to keep normal brain metabolism, and, therefore, the brain is in enhanced anaerobic glycolysis. It is possible that the increased CSF lactate is an epiphenomenon resulting from gliosis and macrophage accumulation in patients in subacute to chronic infarcts. Finally, lactate is increased in the brain, as well as in CSF. These results strongly suggest the existence of a metabolically ischemic state of the brain (increased anaerobic glycolysis with oxygen metabolism maintained) in chronic cerebral infarction or TIA, particularly in some patients who have steno-occlusive vascular lesions. Misery perfusion or mismatch between CBF and CMRO2 has been reported in patients with occluded internal carotid artery, in which up to 20% of patients may experience delayed ischemic events in the ipsilateral cerebral hemisphere. Baron et al4 described in stroke patients with carotid occlusion less decrease in cerebral metabolic rate for glucose than that for oxygen (CMRglc ⬎ CMRO2), indicating enhanced anaerobic glycolysis. Their findings support our results of enhanced anaerobic glycolysis and increased CSF lactate. Correlation coefficiency was much greater between CSF lactate and mCMRO2/CBF than that between CSF lactate and mOEF, indicating that the value for mCMRO2/CBF is related to, but not equal to, mOEF. In an anemic condition, the difference between 2 parameters becomes greater than that in a nonanemic condition, because CMRO2/CBF is the same as OEF ⫻ arterial O2 content. Therefore, absolute value and fraction rate of cerebral oxygen extraction might be different, depending on the hemoglobin level and oxygen saturation of arterial blood. As mentioned in the Results, there were no essential differences in hemoglobin, hematocrit, and arterial gas parameters including oxygen saturation among the groups in our study.
The present results are strongly suggestive in aspect to hemodynamic treatment in patients with chronic brain infarction with increased OEF greater than 0.45 or increased mCMRO2/CBF greater than 7.9 vol% of oxygen, in which the brain becomes metabolically ischemic or anaerobic. However, CMRO2 is not indicative of ischemic metabolism of the brain. In those patients, restoration of the reduced cerebral perfusion pressure might be promising against the stroke recurrence either by medical treatment with vasodilators or with raising systemic blood pressure,16 or by surgical treatment with endarterectomy of steno-occlusive vessels or extraintracranial bypass.12,13 To confirm our findings more specifically, simultaneous determinations of rOEF and rCMRO2/rCBF by PET and regional tissue lactate or adenosine triphosphate by 1H- or 32P magnetic resonance spectroscopy (MRS) are needed to show a direct correlation of increased OEF and lactate. Irie et al17 from our laboratory have reported in multiinfarct dementia that CSF lactate slightly, but significantly increased, which was well correlated with decreased CBF and increased OEF, but not with CMRO2, indicating that cerebral ischemia with impaired aerobic glycolysis plays a role in dementia in multiple infarction. Therefore, we must study the cognitive function in chronic ischemic stroke patients with steno-occlusive changes in carotid or intracranial major arteries. In conclusion, increased OEF and CMRO2/CBF were significantly correlated with increased CSF lactate, indicating the brain to be in a metabolically ischemic state or in enhanced anaerobic glycolysis with oxygen metabolism maintained in patients with chronic ischemic stroke.
References 1. Powers WJ. Cerebral hemodynamics in ischemic cerebrovascular disease. Ann Neurol 1991;29:231-240. 2. Heiss W-D. Application of positron emission tomography to the study of cerebral ischemia. In: Ginsberg MD, Bogousslavsky J, eds. Cerebrovascular disease: Pathophysiology, diagnosis, and management. Malden: Blackwell Science, 1998;761-772.
ISCHEMIC BRAIN METABOLISM IN CHRONIC CEREBROVASCULAR DISEASE 3. Frackowiak RSJ, Pozzilli C, Legg NJ, et al. Regional cerebral oxygen supply and utilization in dementia. A clinical and physiological study with oxygen-15 positron tomography. Brain 1981;104:753-778. 4. Baron JC, Bousser MG, Rey A, et al. Reversal of focal ‘‘misery-perfusion syndrome’’ by extra-intracranial arterial bypass in hemodynamic cerebral ischemia. A case study with 15O positron emission tomography. Stroke 1981;12:454-459. 5. Tsutsumi K, Nagata K. Misery perfusion syndrome at the chronic stage of cerebral infarction (in Japanese with English abstract). Jpn J Stroke 1998;20:489-499. 6. Lenzi GL, Franckowiak RSJ, Jones T. Cerebral oxygen metabolism and blood flow in human cerebral ischemic infarction. J Cereb Blood Flow Metab 1982;2:321-335. 7. Gibbs JM, Frackowiak RSJ, Legg NJ. Regional cerebral blood flow and oxygen metabolism in dementia due to vascular disease. Gerontology 1986;32:84-88, (suppl 1). 8. Yamauchi H, Fukuyama H, Fujimoto N, et al. Significance of low perfusion with increased oxygen extraction fraction in a case of internal carotid artery stenosis. Stroke 1992;23:431-432. 9. North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med 1991;325:445-453. 10. Fujii K, Sadoshima S, Okada Y, et al. Cerebral blood flow and metabolism in normotensive and hypertensive pa-
11.
12.
13.
14.
15.
16.
17.
171
tients with transient neurologic deficits. Stroke 1990;21: 283-290. Fujishima M, Sugi T, Choki J, et al. Cerebrospinal fluid and arterial lactate, pyruvate and acid-base balance in patients with intracranial hemorrhages. Stroke 1975;6:707714. Gibbs JM, Wise RJS, Leenders KL, et al. Evaluation of cerebral perfusion reserve in patients with carotid-artery occlusion. Lancet 1984;1:310-314. Busse O, Hoffmann O. CSF lactate and CT findings in middle cerebral artery infarction. A comparative study. Stroke 1983;14:960-963. Berger JP, Frawer R. Cerebrospinal fluid (CSF) lactate and pyruvate in acute neurological situations. In: Bossart H, Lausanne P, eds. Lactate in acute conditions. Basel: Karger, 1979:115-133. Baron JC, Rougemont D, Soussaline F, et al. Local interrelationships of cerebral oxygen consumption and glucose utilization in normal subjects and in ischemic stroke patients: A positron tomography study. J Cereb Blood Flow Metab 1984;4:140-149. Meyer JS, Judd B, Tawaklna T, et al. Improved cognition after control of risk factors for multi-infarct dementia. JAMA 1986;256:2203-2209. Irie K, Ibayashi S, Fujii K, et al. Relationship between cerebral blood flow, metabolism and cerebrospinal fluid lactate in multi-infarct dementia (in Japanese with English abstract). Jpn J Stroke 1994;16:15-20.
